{"id":"W4414539638","doi":"10.1109/icc52391.2025.11161303","title":"Leveraging MoE-Based Large Language Model for Zero-Shot Multi-Task Semantic Communication","year":2025,"lang":"en","type":"article","venue":"","topic":"Robotics and Automated Systems","field":"Engineering","cited_by":1,"is_retracted":false,"has_abstract":true,"ca_institutions":"University of British Columbia","funders":"","keywords":"Leverage (statistics); Language model; Retraining; Architecture; Generalization; Feature (linguistics); Feature extraction; Wireless","routes":{"ca_aff":true,"ca_fund":false,"ca_venue":false,"about_ca":false,"invisible_to_affiliation_only":false},"retraction":null,"screen":null,"direct_labels":[],"prediction":{"model_version":"codex-gemma-dda1882f352a","candidate_categories":[],"consensus_categories":[],"category_scores_codex":[0.0001689938,0.0001075109,0.0001436081,0.00008670182,0.00008699055,0.00004890954,0.0001641896,0.00006104958,0.000006027069],"category_scores_gemma":[0.00001259865,0.0001041552,0.00005621739,0.0001005456,0.00000622714,0.00004865829,0.00002300296,0.00007545223,0.00001056943],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.00004280162,"about_ca_system_score_gemma":0.00001685401,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.00003511324,"about_ca_topic_score_gemma":0.0000497802,"domain_scores_codex":[0.9994298,0.00001424226,0.0001961771,0.0001047985,0.0000571204,0.0001979003],"domain_scores_gemma":[0.9995117,0.00005104676,0.00001522488,0.0003647783,0.00003178529,0.000025483],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"simulation_or_modeling","study_design_gemma":"simulation_or_modeling","study_design_scores_codex":[0.000001393228,0.00002771289,0.00008791884,0.0002147354,0.00002876278,4.988931e-7,0.0003991387,0.9857391,0.01037661,0.001379325,0.001514968,0.0002298357],"study_design_scores_gemma":[0.0006833237,0.000002344155,0.00006130429,0.00009972922,0.00001511641,2.172998e-7,0.0001546905,0.9947616,0.003788679,0.000073451,0.0002452266,0.0001142924],"study_design_candidate":"simulation_or_modeling","study_design_consensus":"simulation_or_modeling","genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.02518503,0.0004683002,0.9714035,0.0001117245,0.0001057493,0.0002888147,0.00001514242,0.0005408057,0.001880966],"genre_scores_gemma":[0.976431,0.000008670569,0.02153849,0.0001531184,0.000006319191,0.00004611755,0.00006363905,0.00002439097,0.001728257],"genre_candidate":"empirical","genre_consensus":null,"teacher_disagreement_score":0.951246,"threshold_uncertainty_score":0.4247325,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.02715326570425471,"score_gpt":0.2765807636428684,"score_spread":0.2494274979386137,"validation_status":"score_only:v0-immature-baseline","note":"Baseline scores from an immature model (maturity gate not passed). Scores rank; they never assert a category."}}